The first single-stranded DNA virus targeting Pectobacterium belongs to the family Microviridae and demonstrates a broad host range to Pectobacterium brasiliense soft rot pathogens

TL;DR

This paper describes the first single-stranded DNA virus that targets Pectobacterium, a harmful bacteria causing plant diseases, and shows it could be a useful biocontrol agent.

Contribution

The discovery of the first single-stranded DNA phage from the Microviridae family targeting Pectobacterium.

Findings

Phage Mimer infects a broad range of Pectobacterium brasiliense isolates.

Phage Mimer has a latent period of 65 minutes and a burst size of about 79 virions per cell.

Phage Mimer is proposed to belong to a new genus within the subfamily Bullavirinae.

Abstract

Pectobacterium brasiliense (Pbr) is known to be one of the most virulent Pectobacterium species and is generally widely distributed across the globe, especially known for causing soft rot in potato-tubers and black leg in potato-plants. Currently no treatment mechanism exists, and many studies have focused on alternative treatment approaches such as biocontrol agents. Several studies have used phages targeting Pectobacterium species as effective biocontrol agents. This study is the first description of a single-stranded DNA phage belonging to the family Microviridae that targets Pectobacterium. The novel Pectobacterium phage Mimer is proposed to belong to a new genus within the subfamily Bullavirinae in the family Microviridae. Phage Mimer has a genome size of 5879 nt with twelve predicted gene products. Seven out of twelve gene products could be assigned with a function based on either amino acid sequence or structural similarity. Gene synteny and phylogenetic analyses suggest that phage Mimer is part of the subfamily Bullavirinae. Phage Mimer proved to infect a broad range of Pbr isolates but showed a poor adsorption rate as only 17% of phage particles adsorbed within 10 min on the isolation host. Growth kinetics showed phage Mimer to have a latent period of 65 min and an average burst size of approximately 79 virions per cell. Phage Mimer is the first ssDNA phage targeting Pectobacterium and could be a promising biocontrol agent with great therapeutic potential, based on both the small genome size and the well-known genome architecture of model phage phiX174.

The online version contains supplementary material available at 10.1007/s00705-026-06543-2.

Figures

Intergenomic similarity score, gene synteny of genus representatives within the subfamily *Bullavirinae* and phage Mimer and selected protein structures. *Bullavirinae* genus representatives are as follows; *Sinsheimervirus* (phage phiX174) (blue), *Gequatrovirus* (phage G4) (purple), *Alphatrevirus* (phage alpha3) (light pink), proposed genus “*Mimervirus*” (phage Mimer) (green). A Intergenomic similarity score between all phages present in the subfamily *Bullavirinae*: Enterobacteria phage alpha3 (acc. no.: [NC_001330](NC_001330)), Enterobacteria phage phiX174 (acc. no. [NC_001422.1](NC_0014

Phylogenetic analyses using maximum likelihood test of conserved phage proteins within the family *Microviridae,* using predicted protein structures by AlphaFold2 [[37](#CR37)] (A, C, E, G) and aa sequences (B, D, F, H). Branch scale for phylogenetic analysis using aa sequences (B, D, F, H) indicates number of substitutions per site, whereas branch scale for phylogenetic analysis using secondary structures (A, C, E, H) represents structural distances derived from Foldseek scores, indicating structural divergence between proteins. Branch values are noted in black text for all trees, bootstrap v

Morphology and growth kinetics of phage Mimer. a Transmission electron microscopy of phage Mimer shows Bullavirus morphology. White arrows indicate capsid spikes. b Growth kinetics of phage Mimer. The PFU/ml is given on the Y-axis and time in minutes is given on the X-axis. Phage addition to cells at timepoint −5 min. The latent period (defined from adsorption to end of rise period) and stable plateau are annotated on the graph. For each time point the PFU/ml is shown with SD (standard deviation) based on triplicates

Full text

Introduction

Pectobacterium* brasiliense* (Pbr) is part of the group of pectinolytic bacterial plant pathogens: soft rot Pectobacteriaceae (SRP) within the genus Pectobacterium. Pectobacterium is known to cause disease in a wide variety of plants and especially in potatoes, in which it is known to be one of the most economically important plant diseases [14, 49]. Pbr is known to be widely distributed and considered as one of the most virulent species within the genera of Pectobacterium, causing severe losses of potatoes [54]. Control measurements of SRP’s are mainly based on disease management and seed certification systems, as no cultivars have been bred towards resistance and no chemical control measurements can be used due to their adverse effects (such as soil and water contamination, copper-resistant strains etc.) [28, 34, 65]. An increased focus has been directed towards alternative treatment mechanisms such as the use of biocontrol agents like bacteria and bacteriophages (phages) [14]. Since 2012, when the first complete genome sequences of lytic phages targeting SRP were published, an increased number of studies have focused on lytic phages as possible biocontrol agents [8],A. B. [10, 18, 55, 68]. Out of 304 sequenced SRP targeting phages (complete genomes of Pectobacterium phages, the nucleotide database, filtered by viruses, in the NCBI database 27-08-2025), no lytic single stranded DNA (ssDNA) phages have been described [61]. The family Microviridae are non-enveloped ssDNA phages with icosahedral capsids and small genome sizes of 5.3–6.1 kb [40]. Even though no microviruses have yet been considered as therapeutic agents, a renewed focus in recent years has been directed towards their therapeutic potential. Their small genome size and simple genome architecture have been emphasized as promising features in a therapeutic aspect, where their small genome size enables the potential of genetic engineering [40, 59]. Recent research has demonstrated the possibility of genetic engineering where Artificial Intelligence (AI) has been used to design viable phage genomes based on the widely used model organism Escherichia phage phiX174, further supporting the therapeutic potential of microviruses [38].

ssDNA viruses within the family Microviridae are still underrepresented in terms of both, number of phage isolates and phage genera [40]. As the majority of isolation procedures have been optimized for double stranded DNA (dsDNA) phages, most ssDNA phage genomes in current databases, belonging to the family Microviridae, arrive from metagenomic studies [15, 39]. Metagenomic studies have proven ssDNA phages within the family Microviridae to be highly abundant in a wide range of environments, including marine systems, freshwater lakes, the human gut and wastewater samples [39, 44]. The family Microviridae includes two subfamilies: Bullavirinae and Gokushovirinae. The two subfamilies differ in both morphology, genome organization and host range, where the subfamily Gokushovirinae infects obligate intracellular parasitic bacteria and the subfamily Bullavirinae infects enterobacteria [15]. Both morphology and gene synteny are shared within the subfamily Bullavirinae, to which phage phiX174 belongs [15, 45]. Microviruses have only been isolated on 11 genera of bacteria, and according to the current International Committee on Taxonomy of Viruses (ICTV) classification, phages being part of the subfamily Bullavirinae have only been isolated on two bacterial genera Escherichia and Salmonella [40]. According to the ICTV (based on the 2023 release: https://ictv.global) the three genera present within the subfamily Bullavirinae are Alphatrevirus, with 10 species, Gequatrovirus, with three species and Sinsheimervirus with only one species (Sinsheimervirus phiX174).

Here we isolate and investigate the phenotype and genetics of Pectobacterium phage Mimer, a member of a proposed new genus termed Mimervirus, within the Microviridae subfamily Bullavirinae targeting Pbr. Pectobacterium phage Mimer has very limited nucleotide similarity with other phages but does share genome length, aminoacid (aa) similarity, gene synteny and morphology with species within the subfamily Bullavirinae.

Materials and methods

Isolation and purification

Phage Mimer was isolated from an organic waste sample as described elsewhere [43]. Briefly, the organic waste sample was centrifuged at 10,000 × g for 10 min at 4 °C and filtered through a 0.45 µm PVDF syringe filter (Fisherbrand^TM^) prior to use for standard double agar overlay [41]. The bacterial host Pbr J47 used for isolation was obtained in a previous study [56]. The bacterial host was grown in LB liquid media at 28 °C overnight and used for double agar overlay for phage isolation. Standard double agar overlay was done with 100 µl of overnight host bacterial culture and 200 µl of pretreated organic waste sample into 4 ml LB medium supplemented with 0.4% agarose and 10 mM CaCl_2_ and 10 mM MgCl_2_. Phages were purified 3 times using the double agar overlay as described above, plaques were picked using a 500 µl pipette tip and stored in SM buffer (100 mM NaCl, 8 mM MgSO_4_·7H_2_O, 50 mM Tris-Cl pH 7.5). Purified phages were used to prepare high titer amplifications using 100 µl of the purified phage stock in SM buffer mixed with 100 µl of overnight culture of Pbr J47 in 10 ml LB medium (supplemented with 10 mM CaCl_2_ and 10 mM MgCl_2_). The amplification was incubated shaking at 28°C overnight and centrifuged at 10,000 × g for 10 min at 4 °C and filtered through a 0.45 µm PVDF syringe filter.

DNA extraction and sequencing

DNA extraction was performed on phage lysates as described elsewhere (A. [11]). Briefly, 5 U of DNAse and RNase (A&A Biotechnology) were added to 200 µl of phage lysate (with a titer of 1×10^8^ Plaque Forming Units (PFU)/ml) following incubation for 1 h at 37 °C. After incubation, 20 µl of 50 mM EDTA and SDS to a final concentration of 0.1% was added to the lysate. 10 µl of Proteinase K (A&A Biotechnology) [20 mg/ml] was then added and the lysate was incubated for 1 h at 55 °C, following 10 min at 70 °C. DNA was purified using DNA Clean & Concentrator™−5 (Zymo Research) according to manufacturer’s protocol (DNA was eluted in 20 µl elution buffer). Sequencing library was built using NEBNext^®^ Ultra™ II DNA Library Prep Kit for Illumina^®^ (New England Biolabs) according to manufacturer’s protocol and the indexed DNA was sequenced with NextSeq500 platform using the Illumina Mid Output Kit v2 (300 cycles), as previously described [26].

Assembly and annotation

The quality of the phage sequencing reads was assessed using FastQC v. 01.11.9 [2]. Low-quality bases and adaptors were removed with Trimmomatic v. 0.39 [7]. For assembly, SPAdes v.3.14.1 was used [3]. The genome assembly was manually corrected to remove terminal redundant duplicates using CLC Genomics Workbench v. 22.0 (QIAGEN, United States). The assembly was cross verified using CLC genomic workbench. Briefly, sequence trimming and assembly was performed using CLC Genomic Workbench V22 using standard settings (QIAGEN, United States). Because of the low nucleotide similarity to previous isolated phages and the widespread nature of overlapping genes in Microviridae, traditional automatic gene calling and annotation was insufficient. Instead, automatic gene calling from GeneMark [6] and Glimmer [20] was manually curated using DNA master as described before [36, 57]. The ORFs were manually annotated using both HHPred [63] against the four databases (pFam-A_V37, SCOP70_2.08, NCBI_Conserved_domains(CD)_V3.19, PDB_mmCIF70) and blastp against the NCBI nonredundant databases.

Comparative genomics and phylogenetic analysis

Comparative genomics of phage Mimer and one representative species isolate within each genus in the subfamily Bullavirinae: Enterobacteria phage phiX174 (acc. no. NC_001422.1), Enterobacteria phage alpha3 (acc. no. NC_001330) and Escherichia phage G4 (acc. no. NC_001420), was carried out using clinker v. 0.0.28 [27]. Likewise, the same species isolates were used to calculate the intergenomic similarity score between the four representative species. The intergenomic similarity score is based on nucleotide similarity as well as genome length comparison using VIRIDIC [51]. The same species representatives for each genus in the subfamily Bullavirinae were also used for phylogenetic analysis (phage phiX174, alpha3 and G4). Conserved Microviridae proteins (internal scaffolding protein, major capsid protein, pilot protein and rep protein) were found in each phage using blastp on the chosen protein in phage Mimer against protein sequences for each genus representative [61]. AlphaFold2 [37] was used to predict secondary structure for selected proteins. Foldseek was used to search for similarities between homolog proteins in phage Mimer and phiX174 which did not share any aa sequence similarity. e-value was calculated as qTMscore+tTMscore)/2, qTMscore =TMscore normalized by query length, tTMscore = TMscore normalized by target [66]. Protein alignments were visualized using PyMoL [62]. For the phylogenetic analysis we also included Bdellovibrio phage phiMH2K (acc. no. AF306496.1) which is part of the Microviridae subfamily Gokushovirinae. Phage proteins from each of the five phages within each conserved protein group were then used for phylogenetic analysis in Phylogeny.fr [21]. Phylogeny.fr uses MUSCLE [22] for multiple alignment, Gblocks [12] for automatic alignment curation, PhyML [31] for tree building using the maximum likelihood method and TreeDyn [16] for tree drawing. Afterwards, ITOL [46] was used for tree visualization and annotation. AlphaFold2 [37] was used to predict secondary structure for all conserved proteins within each phage representatives, and Foldtree [50] was then used to create phylogenetic trees for protein structures using Foldseek [66]. Foldtree uses Foldseek to align protein structures and generate distance metrices (on the pairwise protein alignment) for phylogenetic tree construction. Following this, ITOL [46] was again used for tree visualization and annotation.

TEM imaging

PEG precipitation of the phage lysate was carried out prior to purification to concentrate phage particles. In brief, to prepare a high-volume phage amplification, 100 µl of phage lysate (corresponding to ~ 7.3 × 10^3^ PFU) and 100 µl of bacterial host overnight culture (corresponding to ~ 3 × 10^7^ CFU) was added to 200 ml of LB broth supplemented with 10 mM CaCl_2_ and 10 mM MgCl_2_ and stirred overnight at 28 °C at 200 rpm. The amplification was then centrifuged at 10,000 × g for 10 min at 4 °C and the phage lysate was poured into a sterile flask. Afterwards, 11.7 g of NaCl (Sigma Aldrich) and 20 g of PEG 8000 (Sigma Aldrich) were added to the phage lysate of 200 ml and stirred overnight at 5 °C at 200 rpm. The phage lysate was then centrifuged at 10,000 × g for 10 min at 4 °C. The supernatant was poured off and the PEG pellet was resuspended with 10 ml of SM buffer and stirred overnight at 5 °C at 200 rpm. The following day the PEG solution was again centrifuged at 10,000 × g for 10 min at 4 °C and the supernatant was filtered using 0.22 µm PVDF syringe filter (Fisherbrand^TM^) and kept for further use. CsCl gradient ultracentrifugation was then used with the PEG precipitation as input to obtain highly concentrated and purified phages for morphological analysis, as described earlier [29]. Morphological analysis on phage particles was conducted as described before [58]. Briefly, 10 µl of purified phages were pipetted on 100-mesh copper grids coated with carbon and adsorption was allowed for 20 min. Subsequently, the grids were washed twice with 10 µl deionized water and negatively stained with 10 µl 2% uranyl acetate. Electron micrographs were generated on a Talos L120C transmission electron microscope (ThermoFisher Scientific, Eindhoven, The Netherlands) using a 4 k x 4 k Ceta camera (ThermoFisher Scientific, Eindhoven, The Netherlands) set to an acceleration voltage of 80 kV.

Phage size was measured using the TEM analysis software Velox v3.9.0 (ThermoFisher Scientific, Eindhoven, Netherlands) and the mean and standard deviation of 21 phage particles were calculated. The TEM images were manually improved using GIMP v2.10.32 for contrast and brightness adjustment.

Phage growth kinetics

A single burst-size experiment was performed to estimate both latent period and burst size for phage Mimer, as described elsewhere [17]. Pbr J47 was grown at 28 °C in LB broth supplemented with 10 mM CaCl_2_and 10 mM MgCl_2_ at 200 rpm to OD_600_ 0.3, corresponding to approx. 6.67 × 10^7^ CFU/ml (a relative low cell concentration was used to ensure that the cells are still in exponential growth, as Pbr J47 leaves exponential phase early). The cells were then mixed with phages at a multiplicity of infection (MOI) of 0.75 and incubated at 28 °C at 200 rpm for 10 min, to allow adsorption. The mixture was then centrifuged for 6000 × g for 5 min. The supernatant, containing the unabsorbed phages, and the pellet, containing the adsorbed phages, was diluted and used for plaque assays in triplicates to estimate adsorption rate. The cell-phage mixture (pellet) was diluted 1 × 10^−4^ in preheated LB broth supplemented with 10 mM CaCl_2_ and 10 mM MgCl_2_, incubated at 28 °C at 100 rpm, and sampled in triplicates every 10 min, diluted and used for plaque assays for 3 hours to determine latent period and burst size. Standard double agar overlay was used for all plaque assays during the setup, as described elsewhere [41]. In brief, 100 µl of overnight culture of Pbr J47 (corresponding to approx. 3 × 10^7^ CFU) and 100 µl of the cell-phage sample was added to the overlay agar of 4 ml LB medium supplemented with 0.4% agarose and 10 mM CaCl_2_ and 10 mM MgCl_2_. Data processing was carried out using RStudio version 2023.6.1.524 [60], using the data visualization package ggplot2 [67]. The Poisson distribution was used to estimate the probability of multiple infections in one cell based on the MOI_actual_ [1].

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p\left(r\right)=\frac{{n}^{r}{\mathrm{e}}^{-n}}{\mathrm{r}!}$$\end{document}

p probability of r incidences r number of phages adsorbed per bacteria n MOI actual

Host range analysis

To determine the initial host range of phage Mimer, the phage lysate was spotted on isolates from several species of Pectobacterium. The initial spot test only showed clearance (lytic activity) on Pbr isolates. All bacterial isolates used in this study were isolated in a former study [56]. All strains were grown in LB broth overnight at 28 °C at 200 rpm. In all cases where the initial spot test showed clearance, both phage Mimer as well as a filtrate of the original host Pbr J47 (to estimate any prophage activity) was used to determine efficiency of plaquing (EOP) on all respective hosts. To prepare the filtrate of Pbr J47 an overnight culture was centrifuged at 10,000 × g for 10 min at 4 °C and filtered using a 0.45 µm PVDF syringe filter (Fisherbrand^TM^). For the initial spot test as well as the EOP, the titer of phage Mimer was 7.3 × 10^5^ PFU/ml. Standard double agar overlay was done with 100 µl of overnight culture of each respective host into 4 ml LB medium supplemented with 0.4% agarose and 10 mM CaCl_2_ and 10 mM MgCl_2_. Dilutions of either phage Mimer or a filtrate of Pbr J47 (prophage control) were diluted and spotted in triplicates on each host.

Results

Phage Mimer shares genome synteny but little nucleotide similarity with members of the subfamily Bullavirinae

Phage Mimer is a ssDNA phage and has a genome size of 5879 nt. When using megablast to search the nr/nt database we only obtained one hit to Mimer (query cover of 1%), which was to Enterobacteria phage phiX174 (acc. no. EF380015.1) due to a match of 38 bp with 1 gap [61]. Phage Mimer only showed a limited intergenomic nucleotide similarity score to the genus representative phages phiX174, G4 and alpha3 within the Microviridae subfamily Bullavirinae (fig. 1A), with an intergenomic similarity score of 18.9 to Enterobacteria phage phiX174 (acc. no. NC_001422.1) and ~14 to both Enterobacteria phages alpha3 (acc. no. NC_001330) and G4 (acc. no. NC_001420). Phage Mimer thus represents a phage belonging to a new genus, as the intergenomic similarity score to all genus representatives within the subfamily Bullavirinae is <70 [64].Fig. 1. Intergenomic similarity score, gene synteny of genus representatives within the subfamily Bullavirinae and phage Mimer and selected protein structures. Bullavirinae genus representatives are as follows; Sinsheimervirus (phage phiX174) (blue), Gequatrovirus (phage G4) (purple), Alphatrevirus (phage alpha3) (light pink), proposed genus “Mimervirus” (phage Mimer) (green). A Intergenomic similarity score between all phages present in the subfamily Bullavirinae: Enterobacteria phage alpha3 (acc. no.: NC_001330), Enterobacteria phage phiX174 (acc. no. NC_001422.1), Enterobacteria phage G4 (acc. no. NC_001420) and phage Mimer (acc. no. ON872163.1). The intergenomic similarity score was calculated using VIRIDIC [51]. b Genome comparison with all phage representatives (as in a) from each genus within the Microviridae subfamily Bullavirinae. All genes are assigned with a function based on the genome of Enterobacteria phage phiX174; gpF (purple), gpG (pink), gpH (red), gpA (orange), gpA* (light orange, only present in phage phiX174), gpB (yellow), gpK (light green), gpC (green), gpD (turquoise), gpE (light blue), gpJ (dark blue). Furthermore, gpG (purple) and gpH (dark red) are assigned with a function based on secondary protein structure hits and unknown (grey) only present in phage Mimer. c Alignment of protein structure predicted by AlphaFold2 of GpG and GpH in phage Mimer (green) and phage phiX174 (blue). Foldseek alignment score resulted in e-value 3.59 × 10^−12^ for GpG and 2.17 × 10^−1^ for GpH

Phage Mimer encodes twelve putative genes (fig. 1B, suppl. table 1), for which a function could be assigned for seven of the twelve genes using aa sequence similarity with blastp and subsequently HHpred. When using blastp (the nr database), phage protein hits could only be found for five out of twelve gene products (with e-values < 0.5) [61]. To elucidate the function of the rest of the genes, HHpred [63] (versus the pFam-A_V37, SCOP70_2.08, NCBI_Conserved_ domains(CD)_V3.19, PDB_mmCIF70 databases) resulted in phage protein hits for two out of the seven gene products used as input (with probability score >95%) (suppl. table 1).

Comparison of the genome architecture revealed strict conservation in gene placement for gpG (major spike protein) and gpH (minor spike protein) across all four phages (fig. 1B), despite the absence of aa sequence similarity (suppl. table 1) in phage Mimer. To investigate whether the two genes in phage Mimer (gpG and gpH) not only shared gene placement but also protein structure with their homologs in phiX174, we used AlphaFold2. Protein structures were predicted for all gene products using AlphaFold2 for phage Mimer and phiX174 [37]. Foldseek was used to investigate whether the two predicted proteins from each phage, GpG and GpH, shared protein structure without any aa similarity (fig. 1C) [66]. The Foldseek search using GpG in phage Mimer as input versus all proteins in phage phiX174 resulted in a match to GpG in phiX174 with an e-value of 3.59 × 10^−12^ and indicated shared protein structure (fig. 1C). The Foldseek search using GpH in phage Mimer as input versus all proteins in phage phiX174 resulted in a match to GpH in phiX174 with an e-value of 2.17 × 10^−1^, which indicates a somewhat shared structure (fig. 1C). For all other genes within the genome of phage Mimer as well, the gene synteny appeared well conserved when comparing the genome architecture with the three genus representatives within the subfamily Bullavirinae (fig. 1B).

Phage Mimer represents a new genus within the subfamily Bullavirinae

To investigate the phylogenetic relationships among phage Mimer and the genera within the subfamily Bullavirinae we analysed conserved phage proteins found within the family Microviridae [15]. For the analysis we used the three genus representatives within the subfamily Bullavirinae; phage phiX174, G4 and alpha3, and also included phage phiMH2K, which is part of the Microviridae subfamily Gokushovirinae, as an outgroup. Conserved phage proteins used in the analyses were GpF (internal scaffolding protein), GpB (major coat protein), GpH (DNA pilot protein) and GpA (replication protein). As several gene products in phage Mimer seem to have little or no nucleotide or aa sequence similarity with the four representatives from each genus; phage phiX174, G4 and alpha3 (suppl. table 1, fig. 1B), the phylogenetic analyses were done using both aa sequences and protein structure predictions (fig. 2). The two different approaches differed for each protein for three out of four analyses. As the structural as well as aa sequence similarity within each group of proteins, found in the subfamily Bullavirinae, varied, we did expect different output of the phylogenetic analyses (suppl. fig 1, fig. 2).Fig. 2. Phylogenetic analyses using maximum likelihood test of conserved phage proteins within the family Microviridae, using predicted protein structures by AlphaFold2 [37] (A, C, E, G) and aa sequences (B, D, F, H). Branch scale for phylogenetic analysis using aa sequences (B, D, F, H) indicates number of substitutions per site, whereas branch scale for phylogenetic analysis using secondary structures (A, C, E, H) represents structural distances derived from Foldseek scores, indicating structural divergence between proteins. Branch values are noted in black text for all trees, bootstrap values are noted in bold brown text for aa trees (B, D, F, H). Conserved phage proteins in phage Mimer were used to find the best protein homologs, using blastp [61], in phage phiX174 (acc. no. NC_001422.1), alpha3 (acc. no.: NC_001330) and G4 (acc. no. NC_001420). Conserved phage proteins, corresponding to the four proteins in phage Mimer, were likewise included for phage phiMH2K (acc. no. AF306496.1) and used as an outgroup in all trees. A Phylogenetic analysis using predicted protein structures of GpF (major coat protein) in phage Mimer as well as best protein homologs in phage phiX174, alpha3, G4 and corresponding protein Vp1 in phage phiMH2K. B Phylogenetic analysis using aa sequences of all proteins described in A. C Phylogenetic analysis using predicted protein structures of GpB (internal scaffolding protein) in phage Mimer as well as best protein homologs as described in phage phiX174, alpha3, G4 and corresponding protein Vp3 in phage phiMH2K. D Phylogenetic analysis using aa sequences of all proteins described in C. E Phylogenetic analysis using predicted protein structures of GpH (DNA pilot protein) in phage Mimer as well as best protein homologs as described in phage phiX174, alpha3, G4 and corresponding protein Vp2 in phage phiMH2K. F Phylogenetic analysis using aa sequences of all proteins described in E. G Phylogenetic analysis using predicted protein structures of GpA (replication protein) in phage Mimer as well as best protein homologs as described in phage phiX174, alpha3, G4 and corresponding protein Vp4 in phage phiMH2K. H Phylogenetic analysis using aa sequences of all proteins described in G

The phylogenetic analysis based on structure of the four conserved proteins did not result in the same relationship for all proteins. The phylogenetic analysis using protein structure of GpF (fig. 2A) resulted in phage G4 as an outgroup within the subfamily Bullavirinae, and phage alpha3 as an outgroup within the subfamily using GpA (fig. 2G). Phage Mimer appeared as an outgroup within the subfamily in two of the four analyses, using both GpB and GpH (fig. 2C).

The phylogenetic analysis based on aa sequences of the four conserved proteins indicated that phage Mimer is part of the subfamily Bullavirinae when using phage phiMH2K as an outgroup (fig. 2B). The phylogenetic relationship among the four phages (phage phiX174, G4, alpha3 and Mimer) differ for each protein, and tend to have low statistical support, which is not surprising for small divergent proteins [13]. The phylogenetic analysis using GpF resulted in two subgroups in which phage Mimer and G4, and phage alpha3 and phiX174 grouped together, respectively (fig. 2B). Phage Mimer and alpha3 showed the closest phylogenetic relationship based on GpB and GpH (fig. 2D). For the phylogenetic analysis based on GpA, phage phiX174 and G4 showed to be closest related (fig. 2H). The phylogenetic analysis using the aa sequence and predicted structure of GpA resulted in the same relationship among all phages (fig. 2G).

Phage Mimer displays Bullavirus morphology and shows low adsorption rate to host

Phage Mimer displays a typical Bullavirus-like structure (fig. 3A), which is characterized by an icosahedral shell harbouring prominent spike proteins (arrows at fig. 3A) at each of the five-fold axes of symmetry of the capsid [23]. Size measurements revealed a virion diameter of 28.1 ± 2.3 nm.Fig. 3. Morphology and growth kinetics of phage Mimer. a Transmission electron microscopy of phage Mimer shows Bullavirus morphology. White arrows indicate capsid spikes. b Growth kinetics of phage Mimer. The PFU/ml is given on the Y-axis and time in minutes is given on the X-axis. Phage addition to cells at timepoint −5 min. The latent period (defined from adsorption to end of rise period) and stable plateau are annotated on the graph. For each time point the PFU/ml is shown with SD (standard deviation) based on triplicates

The burst size experiment of phage Mimer showed a latent period (defined as the time from adsorption to end of rise period) of 65 min and an average burst size of 79 virions per cell (fig. 3B, suppl. table 2). Phage Mimer showed a low adsorption rate of only 17% after 10 min (suppl. table 2), measured by the fraction of adsorbed phages present in the pellet and the total amount of phages (calculated as the fraction of adsorbed phages in the pellet together with the fraction of unadsorbed phages present in the supernatant). As a MOI_input_ (multiplicity of infection = ratio of phage to host) of 0.63 was used, an explanation for this difference could be due to multiple infection in some cells. To estimate the possibility of multiple infections in some cells we used the Poisson distribution [1]. Based on the very low adsorption rate, we calculated the MOI_actual_ to be 0.105 compared to the MOI_input_ of 0.75 (suppl. table 2). A MOI_actual_ can be used to get a more exact number of the possibility of multiple infected cells when using the Poisson distribution according to [1].

\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p\left(2\right)=\frac{{0.105}^{2}{\mathrm{e}}^{-0.105}}{2!}=0.005$$\end{document}

According to the Poisson distribution we expected less than 1% (0.005) of multiple adsorbed cells when using the MOI_actual_ of 0.105.

Host range experiments showed phage Mimer to infect a broad range of P. brasiliense isolates

To investigate the host range potential of phage Mimer we used 59 isolates from several species of Pectobacterium: P. brasiliense, P. polaris, P. parmentieri, P. punjabense and P. atrosepticum in an initial spot test (Table 1, Supplementary table 3). Efficiency of plating (EOP) was tested for all isolates in which phage Mimer was able to form clearance (lytic activity) in the initial spot test. Phage Mimer did infect a broad range of Pbr isolates (Table 1). To test for any prophage activity from the phage host of phage Mimer (Pbr J47), a prophage control (filtered ON host culture) was included in the host range experiment (Table 1). The prophage control did result in plaque formation for eight out of thirteen hosts in which phage Mimer was able to infect, which showed that the presence of a prophage in the original host was able to target several of the same bacterial strains as phage Mimer.Table 1.Host range of phage Mimer. Phage Mimer was tested on isolates from several species of Pectobacterium in an initial spot test. Efficiency of plating (EOP) was tested for all bacterial host in which phage Mimer showed clearance in the initial spot test (Pbr isolates). Isolation host is noted as ** (Pbr J47). EOP was calculated as titer on test hosts divided by titer on the isolation host. Titer of phage Mimer (on isolation host) was 7.3 × 10^5^ PFU/ml. To test for any prophage activity of the original host a prophage control (filtrate of an overnight culture of the host of phage Mimer, Pbr J47), was tested on all bacterial host where phage Mimer showed clearance in the initial spot test. EOP of the filtrate of Pbr J47 was calculated as the titer on the host divided by the titer of phage Mimer on the isolation host Pbr J47. GenBank acc. numbers (16S) for all P. versatile, P. polaris, P. parmentieri, P. punjabense and P. atrosepticum isolates used in the initial spot test which did not show any clearance can be found in supplementary table 3Bacterial IDTaxonGenBank acc. no. (16S rRNA)Mimer EOPPbr. J47 (prophage con.) EOPJ13P. brasiliensePP5875721.73-J20P. brasiliensePP587577--J21P. brasiliensePP587578--J24P. brasiliensePP5875803.150.01J25P. brasiliensePP5875812.230.01J26P. brasiliensePP5875824.560.01J31P. brasiliensePP5875872.64-J32P. brasiliensePP5875888.220.01J34P. brasiliensePP5875903.420.004J36P. brasiliensePP5875923.560.02J37P. brasiliensePP5875930.05-J41P. brasiliensePP5875971.23-J44P. brasiliensePP5876001.450.003J45P. brasiliensePP587601--J47**P. brasiliensePP5876021.00-J52P. brasiliensePP5876050.820.01J59P. brasiliensePP587609-Other (spot1 isolate of P. versatile-test)4 isolates of P. polaris-7 isolates of P. parmentieri-2 isolates of* P. punjabense*-15 solates of* P. atrosepticum*-

Discussion

Despite limited nucleotide similarity with anything present in the nr/nt database when searching the genome of phage Mimer using megablast, both genome length, aa sequence similarities, gene synteny and morphology confirmed phage Mimer to be part of the Microviridae subfamily Bullavirinae. It is well known that phages related in terms of subfamily or even genus may share genome architecture despite very little nucleotide similarity [33]. This was also demonstrated from the genome analyses of phage Mimer, a Pectobacterium phage, and the three genus representatives within the subfamily (all Escherichia phages), between which very limited nucleotide similarity and little aa sequence similarity, but a highly conserved genome architecture could be observed. The conserved gene synteny could be a result of the ssDNA genome, found in the family Microviridae, that may play a role in the virion structure, where it is closely associated with the capsid’s inner surface and new genes may affect this structure [25]. Little is known about the evolution of these viruses, but there is growing evidence that genes may be acquired from different sources within the ecological niche and that species-jumps may drive the evolution rather than horizontal gene transfer [25, 42]. The conserved genome synteny among the four phages may also indicate the importance of gene position in terms of function.

Seven out of twelve genes could be assigned with a function based on either aa sequence similarity or structural similarities to known proteins. There were, however, six gene products which were assigned as hypothetical proteins. Several bioinformatic approaches have used the conserved gene synteny in genetic loci, such as structural modules in phage genomes, to reveal protein functions of unknown proteins [19, 30, 47]. As phage Mimer shares genome synteny with all representatives from each genus in the subfamily Bullavirinae, an indication of the function of some of these hypothetical proteins might be inferred from the gene placement. Two genes of particular interest, due to their placement in the genome, are gp10 (assigned as putative gpE, lysis protein in fig. 2B) and the last gene product gp12 (assigned as hypothetical protein in fig. 2B) (suppl. table 1, fig. 2B).

gp10 (putative gpE) did not show any hits using either blastp with an e-value < 0.5 or HHpred with a probability score > 95. It did however, when using HHpred on structural similarities, show a match with a probability score of 76.95 to part of the YES complex (acc. no. 8G01 in the PDB database) formed by the phage protein GpE (lysis protein) (from phiX174) and two proteins involved in the cell wall synthesis (in E. coli) facilitating cell rupture [4, 53]. Given the placement of gp10 (putative gpE) in the genome of phage Mimer (overlapping with gpD in both phage phiX174 and phage Mimer (fig. 2B)), together with a match to part of the YES complex known to cause cell rupture in E. coli when infected with phage phiX174, indicate that gp10 (putative gpE) in phage Mimer might also be involved in cell lysis.

The gp12 gene (encoding a hypothetical protein) in phage Mimer appears as the last gene product in the genome, which also is the case for gpJ (encoding DNA packaging protein) in phiX174 (as well as in the phages G4 and alpha3). gp12 in phage Mimer did not show any matches either on aa level with e-values < 0.5 using blastp or structural with probability score > 95 using HHpred. It did, however, have a partial match to a thymine-DNA glycosylase (acc. no. 1KEA_A in the PDB database) in E. coli, classified as a DNA hydrolase with a probability score of 79.56. GpJ (DNA packaging protein) in phage phiX174 facilitates DNA packaging by binding to the ssDNA genome and to the internal surface of the capsid [5]. No clear function can be added to the Gp12 (hypothetical protein) in phage Mimer, however both gene placement and a partly match to a DNA hydrolase could indicate that the protein may be a functional substitute for GpJ in phiX174.

When comparing aa sequences of all predicted gene products in phage Mimer with data available in the nr/nt databases it was obvious that some of these proteins did not have any aa similarities with known protein sequences, but with predicted protein structure more gene products were able to be assigned with a function (fig. 1C, suppl. table 2). Based on this knowledge, we used two different approaches to investigate the phylogenetic relationship between phage Mimer and the three genus representatives within the subfamily Bullavirinae, as both aa sequence and predicted protein structure were used in the phylogenetic analyses (fig. 2). Newer methods have increased availability of structure prediction, and a renewed focus has been pointed towards how to use this in phylogenetic analyses [37, 50]. As protein structure is more conserved than the underlying aa sequence, this approach could be particularly relevant in comparing distantly related species [35, 48]. The phylogenetic analysis did however vary for each protein for three out of four analyses but did demonstrate phage Mimer to be closely related to the genus representative within the subfamily Bullavirinae.

The four conserved phage proteins used in the phylogenetic analysis were GpF, GpB, GpH and GpA, which also is used by ICTV when classifying the family Microviridae in the ninth report [15]. The phylogenetic analysis based on aa sequences of all four proteins proved phage Mimer to be part of the subfamily Bullavirinae together with phage phiX174, G4 and alpha3, when using phage phiMH2K as an outgroup. The formation of clades within the subfamily Bullavirinae did vary based on input protein, which also is seen in the report of Cherwa et. al. The four phylogenetic analyses based on the protein structure did also vary in which phages that clustered together within the subfamily Bullavirinae. Only one of the four analyses did result in the same phylogenetic analyses independent of method (fig. 2G, 2).

As the phylogenetic analyses using protein structures as input no bootstrap values are formed, as the model is built upon pairwise structure comparisons, which makes comparison of the two models rather difficult [50]. Another issue which should be considered is whether the protein structure of several of these proteins found across the family Microviridae might be conserved to a certain extent which may make it difficult to use for phylogenetic analyses. No definitive conclusion on the exact phylogenetic placement of phage Mimer can be drawn, but the phylogenetic analyses together with the gene synteny across the subfamily Bullavirinae and phage Mimer, suggest phage Mimer to be part of the Microviridae subfamily Bullavirinae.

In the burst size experiment, we observed a low adsorption rate of phage Mimer to Pbr J47, as only 17% was adsorbed after 10 min of incubation indicating inefficient adsorption. However, longer incubation periods (>10 min) were not tested, and the low adsorption rate might reflect slow adsorption rather than inefficient adsorption. As phage adsorption of microvirus is poorly understood together with the fact that no other studies have experimental data on a microvirus targeting Pectobacterium, it is unknown whether the adsorption rate may be unusual or not [25]. We have, however, in a previous study observed a likewise low adsorption rate for another Pectobacterium phage, in which we discussed whether this could be due to the low cell density used [55]. In the current study we did also use a relatively low cell density (6.67 × 10^7^ CFU/ml) in contrast with other studies using ~ 1 × 10^8^ CFU/ml [9, 24, 52]. The low adsorption rate of phage Mimer could also imply that the phage may be optimized for another host, which could be supported by the fact that phage Mimer produces an up to eight times higher titer on another host (J32) than the isolation host in the host range experiment (table 1). Furthermore, we observed a prophage of the host (table 1) which were able to infect several of the same isolates as phage Mimer. We did however observe a very low EOP in the prophage control in the host range analysis, and phage Mimer to target more broadly than the prophage of the host. We cannot exclude the possibility of the prophage interfering with the host range results of phage Mimer and in relation to the biocontrol perspective of phage Mimer it would be important to investigate whether another host could be used for propagation. Phage Mimer tends to have a latent period of 65 min with an average burst size of 79 virions per cell, which stands in contrast with phage phiX174. Phage phiX174 is one of the fastest growing phages with an eclipse period of 20 min and burst size of approx. 200 virions per cell [40]. These traits together with a small genome size have led to the general conception of why microviruses are being widely distributed across environments and highly abundant (based on metagenomes studies) [39]. However, recent studies identified other microviruses that were isolated from host species other than E. coli, where longer eclipse periods and lower burst sizes have been observed, as in phage Mimer [32, 40, 59]. Phage Mimer has an average burst size of 79 virions per cell. As a relatively high MOI_input_ of 0.63 was used, we may have underestimated the burst size as some cells may be infected by several phages. But based on MOI_actual_ which showed to be lower than the MOI_input_, due to the low adsorption rate, we were able to determine the probability of multiple infected cells to be less than 1% using the Poisson distribution. We do however, when following the Poisson distribution, assume equal receptor distribution across all cells.

Phage Mimer proved to infect broadly within the species of Pbr where it targeted most isolates in the host range experiment. Phage Mimer did not infect any related species, which together with the broad within-species host range could be beneficial in a biocontrol aspect. Storage experiments together with in planta applications could further elaborate on the biocontrol perspective of phage Mimer.

Conclusion

Here we present a new Pectobacterium phage, phage Mimer, targeting P. brasiliense. Phage Mimer shares morphology, genome length and gene synteny with other phages being part of the Microviridae subfamily Bullavirinae, despite very limited nucleotide similarity. Most gene products (seven out of twelve) could be assigned with a function based on either aa sequence similarity or structural similarity. As some gene products of phage Mimer shared no nucleotide or aa sequence similarity to related phages, we utilized phylogenetic analyses using both aa sequences as well as protein structure which placed phage Mimer as part of the Microviridae subfamily Bullavirinae. Phage Mimer showed an adsorption rate of only 17% after 10 min, a latent period of 65 min and an average burst size of 79 virions per cell, which challenges the fast growing microvirus paradigm. To our knowledge this is the first isolated microvirus able to infect Pectobacterium species. The isolation strain did not seem to be the optimal host (based on EOP values and adsorption rate), but phage Mimer showed potential to strongly infect other P. brasiliense isolates. Phage Mimer might be an efficient biocontrol candidate for controlling P. brasiliense, both in relation to the host range but also the small genome size, which have been argued to be beneficial in several therapeutic aspects, and as a novel source of unexplored antibacterial gene products.

Declaration

Lars Hestbjerg Hansen reports financial support was provided by Novo Nordisk Foundation. The authors have no relevant financial or non-financial interests to disclose. Data collection, experimental design, data analysis and writing the draft manuscript was done by Julie S. Pedersen. Funding acquisition and supervision of the study was performed by Alexander B. Carstens, Witold Kot and Lars H. Hansen. Furthermore, Alexander B. Carstens and Lars H. Hansen contributed to the study with data analysis and experimental design. Supervision, experimental design and data analysis were also performed by Gemma Atkinson. Data collection was carried out by Magnus Rothgardt and Anouk Viry, and data analysis by Artyom Egorov. Sample preparation, TEM imaging and data analysis were conducted by Frank Hille and Charles M.A.P. Franz. All authors commented on previous versions of the manuscript and approved the final manuscript.

Supplementary Information

Below is the link to the electronic supplementary material.Supplementary file1 (TIFF 286875 KB)Supplementary file2 (XLSX 13 KB)Supplementary file3 (XLSX 10 KB)Supplementary file4 (XLSX 10 KB)